RAXE Intelligence

Stay ahead of AI threats

Original threat research from RAXE Labs feeds directly into detection signatures. Every advisory, every technique mapping, every new signature makes the platform stronger.

View Advisories Threat Intelligence
Research Output

The research-to-product loop

RAXE Labs discovers threats. Intelligence turns them into enforceable protection. This creates a compounding advantage that grows with every new finding.

1,000+
Detection Signatures
15+
Published Advisories
4
Research Streams
Monthly
Threat Reports
4 Research Streams

Comprehensive threat coverage

S1

Adversarial ML

Model exploitation, jailbreaking, adversarial inputs, unsafe deserialisation, and techniques that abuse ML model behaviour or loading mechanisms.

S2

Agent Security

AI agent tool-use abuse, MCP server vulnerabilities, sandbox escapes, path traversal via agentic workflows, and inter-agent communication attacks.

S3

Supply Chain

Vulnerabilities in AI/ML frameworks, model registries, training pipelines, and third-party dependencies used across the AI software supply chain.

S4

Prompt Injection

Direct and indirect prompt injection, instruction override, web-based injection against AI agents, and techniques to hijack LLM-powered applications.

Open Detection Content

1,000+ YAML-based signatures

Open, auditable, and continuously updated. Every signature is mapped to MITRE ATLAS techniques and OWASP LLM Top 10 categories. 

id: raxe-pi-001
name: Direct System Prompt Override
severity: critical
stream: S1
mitre_atlas: AML.T0051
owasp_llm: LLM01
patterns:
  - "ignore (all |any )?(previous |prior )?instructions"
  - "disregard (your |the )?system prompt"
  - "you are now (a |an )?[\\w]+ (mode|assistant)"
action: block
confidence: 0.95
How Intelligence Connects
Research becomes protection. Every signature, advisory, and technique mapping produced by RAXE Intelligence is deployed automatically to RAXE Gateway and RAXE Sensor. Gateway enforces policy on AI API traffic. Sensor enforces policy at runtime. Intelligence provides the detection content that powers both.

Deploy Intelligence-powered protection

Explore our research or deploy the platform that enforces it.

Get a Demo Start Free
Security Advisories → Threat Intelligence →